SUPPLEMENTARY APPENDIX HOXC4 homeoprotein efficiently expands human hematopoietic stem cells and triggers similar molecular alterations as HOXB4

Céline Auvray,1,2 Andrée Delahaye,1-3 Françoise Pflumio,4 Rima Haddad,5 Sophie Amsellem,6 Ayda Miri-Nezhad,1,2 Loïc Broix,1,2 Azzedine Yacia,1,2 Frédérique Bulle,1,2 Serge Fichelson,1,2 and Isabelle Vigon1,2

1Inserm U1016, Institut Cochin, Paris; 2Université Paris Descartes, Sorbonne Paris Cité; 3Université Paris 13, INSERM U676, Service Histologie, Embryologie et Cytogénétique, Hôpital Jean Verdier, Bondy; 4Inserm U967, Université Paris 7, Commissariat à l’Energie Atomique, Fontenay-aux-Roses; 5INSERM U935, Faculté de Médecine Paris-Sud 11, le Kremlin-Bicêtre; 6Institut Gustave Roussy, Villejuif, France. Citation: Auvray C, Delahaye A, Pflumio F, Haddad R, Amsellem S, Miri-Nezhad A, Broix L, Yacia A, Bulle F, Fichelson S, and Vigon I. HOXC4 homeoprotein efficiently expands human hematopoietic stem cells and triggers similar molecular alterations as HOXB4. Haematologica 2012;97(2):168-178. doi:10.3324/haematol.2011.051235

Online Supplementary Data to Design and Methods Semi-solid cloning assays Human clonogenic progenitors were tested by plating 500 Isolation and immuno-labeling of CD34+ cells cells in 35-mm duplicate dishes (Dutscher, Brumath, France), in Normal cord blood was collected according to institutional 1 mL methylcellulose (StemCell Technologies) supplemented guidelines and after informed consent of the mothers. with 2 U/mL erythropoietin, 10 ng/mL interleukin-3, 15 ng/mL Following Ficoll separation (Lymphoprep, Fresenius Kabi, stem cell factor, 20 ng/mL granulocyte colony-stimulating factor Sèvres, France), CD34+ cells were enriched with the CD34 (Amgen, Thousand Oaks, CA, USA), and 5 ng/mL granulocyte- microBead kit (Miltenyi Biotec, Bergisch-Gladbach, Germany) macrophage colony-stimulating factor (GM-CSF; Schering- according to the manufacturer’s instructions. For isolation of Plough, Heist-op-den-Berg, Belgium). Dishes were incubated at + the most immature hematopoietic progenitor cells, CD34 cells 37°C in a fully humidified atmosphere containing 5% CO2. were incubated for 30 min at 4°C with fluorescein isothio- Burst-forming units-erythrocyte (BFU-E), colony-forming units- cyanate-conjugated antibody to CD34 (clone 581, Beckman granulocyte/macrophage (CFU-GM), and colony-forming units- Coulter, Villepinte, France) and phycoerythrin-conjugated anti- granulocyte/erythrocyte/macrophage/megakaryocyte (CFU- body to CD38 (clone T16, Beckman Coulter). GEMM) were counted on day 15 under an inverted microscope.

Co-cultures Western blots CD34+ cells (1,000 cells/cm2) were co-cultured for 4-5 weeks Cells were rinsed and then lysed in Ripa (Tris-HCl pH8, with 25-Gy pre-irradiated MS-5/GFP, MS-5/HOXB4, or MS- EDTA 4 mM, NaCl 150 mM, Triton 0.5%, SDS 0.2%) in the 5/HOXC4 cells in H5100 long-term culture medium (StemCell presence of a protease inhibitor cocktail (Roche, Rosny-sous- Technologies, Grenoble, France). Total cells and CD34+ cells Bois, France). After denaturing polyacrylamide gel elec- were counted every week to evaluate expansion ratios. A cell trophoresis (SDS-PAGE) migration, proteins were transferred sample was seeded every week into methylcellulose as onto a HybondC Extra membrane (Amersham Bioscience, described below for progenitor assays. For quantification of Munich, Germany). The membrane was blocked for 1 h at long-term culture-initiating cells (LTC-IC), CD34+ cells derived room temperature with 5% (w/v) skimmed milk in Tris- from 4- to 5-week-old primary co-cultures were sorted and buffered saline-0.2% Tween (TBS-T), then hybridized in the plated again at limiting dilutions on pre-irradiated non-trans- same buffer with the I12 rat anti-human HOXB4 monoclonal duced MS-5 cells for 4-5 additional weeks, after which the antibody and/or the N-20 goat anti-human HOXC4 antibody number of LTC-IC was determined by counting the number of (sc-49986, Santa Cruz Biotechnology, Santa Cruz, CA, USA), or colony-containing wells. a mouse monoclonal anti-actin antibody (clone AC-15, Abcam, Paris, France) as a control. Antibodies were diluted 1:1000 and Mouse assays incubated overnight at 4°C. Blots were visualized using a horse- For in vivo experiments, expansion rates of CD34+ cells after 4 radish peroxidase-conjugated secondary goat anti-rat (sc-2006, weeks of co-culture varied from x12 (in the presence of MS- Santa Cruz Biotech) anti-mouse antibodies, or a donkey anti- 5/GFP cells) to x30 (in the presence of MS-5/HOXB4 or /C4 cells). goat antibody (sc-2056, Santa Cruz Biotech). The mean total CD34+ cell numbers obtained from 5,000 initial Chemiluminescence was detected (enhanced chemilumines- CD34+ cells was, therefore, 60,000 with MS-5/GFP and 150,000 cence, Amersham Bioscience, Villejuif, France). Images were with MS-5/HOXB4 or /C4. Considering that about 50% of cells captured using a CCD camera (LAS4000, Fujifilm, Tokyo, were retrieved after cell sorting, a mean of 30,000 and 75,000 Japan) and quantification was performed using Image Reader CD34+ cells, respectively, were injected into animals. Las-4000 software (Fujifilm). Online Supplementary Table S1. List of primers used for RT-qPCR analysis of genes from the tran- scriptomes of CD34+ cells exposed to HOXB4 or HOXC4. Gene Symbols Pubmed accession numbers Sequences 5’ TO 3’ ABCE1 (F) NM_002940.2 GCCAAACCTTGGAAAGTACG ABCE1 (R) TTGTAATTCAGATCCACGGAAA BOP (F) NM_015201.3 TCTGAGCCCGAACTGGAG BOP (R) TGCTGTGGCTGAGAGGAGA BRCA2 (F) NM_000059.3 GCGCGGTTTTTGTCAGCTTA BRCA2 (R) TGGTCCTAAATCTGCTTTGTTGC CASP8 AP2 (F) NM_012115.2 GAAATGAAAGCCCACCTCAA CASP8 AP2 (R) GCTGAATTTGGTAAAAACACATCTT CCND2 (F) NM_001759.2 GGACATCCAACCCTACATGC CCND2 (R) CGCACTTCTGTTCCTCACAG C-MYC (F) NM_002467.4 CACCAGCAGCGACTCTGA C-MYC (R) GATCCAGACTCTGACCTTTTGC DACH1 (F) NM_004392.4 AGCAGTTGGCTATGGAACAAA DACH1 (R) CCGTTTCGTCTCAAACTCAAG DBF4 (F) NM_006716.3 TGGAATATGAAAAGGACACACCT DBF4 (R) CAGTCTTTTTCAGGACACTTGC DHX9 (F) NM_001357.3 CTGCCTGTATCACTGGTC DHX9 (R) CTACACACAGGAACTGAC EZH2 (F) NM_004456.3 TGGTCTCCCCTACAGCAGAA EZH2 (R) CCGTTTCGTCTCAAACTCAAG GNL3 (F) NM_014366.4 CAGCAACGTCTATGACCTCCT GNL3 (R) CTTGAAGATCTGCCCAGTCC HBP1(F) NM_012257.3 AATATACTCAGATGTATCCAGGGAAAG HBP1 (R) TTCCACCTGTCACCAAGGA HDAC2 (F) NM_001527.2 CAGATCGTGTAATGACGGTATCA HDAC2 (R) CCTTTTCCAGCACCAATATCC HBG1 (F) NM_000184.2 TGGATCCTGAGAACTTCAAGC HBG1 (R) CACTGGCCACTCCAGTCAC HBA1(F) NM_000558.3 GACCCGGTCAACTTCAAGC HBA1 (R) AGAAGCCAGGAACTTGTCCA HNRNPR (F) NM_001102398.1 TGGAAAACTCGAAAGAGTAAAGAAG HNRNPR (R) CCATTCATTTCATCCATAGCC HSP90AB1 (F) NM_007355.2 AGCCTACGTTGCTCACTATTACG HSP90AB1 (R) GAAAGGCAAAAGTCTCCACCT IFITM1 (F) NM_003641.2 CACGCAGAAAACCACACTTC IFITM1 (R) TGTTCCTCCTTGTGCATCTTC IGFBP2 (F) NM_000597.2 GGTGGCAAGCATCACCTT IGFBP2 (R) TCCTGTTGGCAGGGAGTC IKZF (F) NM_006060.2 CCTTCCGGGCACACTGTA IKZF (R) TCTCTCTGATCCTATCTTGCACA NOL5A (F) NM_006392.2 AATTCCACAGCATCGTTCG NOL5A (R) GCGGAGGTCCTCATGAAC PBXIP1 (F) NM_020524.2 TCAGGGACCTCAGCAACTATG PBXIP1 (R) CTCCACTGGCAGGCTCTC PPAN (F) NM_020230.4 CCATCAACGTGCACAAGG PPAN (R) GAGTCGGGGTTGTAGTCGAT PUM1 (F) NM_014676.2 GTCCTGTCATTGGCACTACAGAT PUM1 (R) CCCGAACCATCTCATTCTG PUM2 (F) NM_015317.1 TGATCTGTGGATGAAATGAATC PUM2 (R) TTTACAGCATTCCATTTGGTG RAD51 (F) NM_002875.3 CAAAGGCGGTCAGAGATCAT RAD51 (R) TGATAGATCCAGTCTCAATTCCAC RCF3 (F) NM_181558.2 GGGACGGCTGGACTATCA RCF3 (R) GGAAAGTCACCACACTGCAC TP53 (F) NM_000546.3 AGGCCTTGGAACTCAAGGAT TP53 (R) CCCTTTTTGGACTTCAGGTG YPEL5 (F) NM_016061.1 CACATCTCCACTCGTTTCACAGG YPEL5 (R) CGCTGTACTGCAGGTTAACTACC Online Supplementary Table S2. List of genes up- or down-regulated after CD34+ cell exposure to HOXB4 and to HOXC4 (fold change threshold = 0.5; false discovery rate, i.e. corrected P-value = 5%). A total of 422 genes were up-regulated and 167 genes were down-regulated (see Table S2 xlsx).

Online Supplementary Table S3. Genes involved in the cell cycle, and whose expression was deregulated in hematopoietic cells after 24 h exposure to HOXB4/C4. Positively regulated Negatively regulated Cyclins: A2, D2, E2 Cyclin D binding protein 1 (CCNDBP1) CDCs: 2, 6, 20 CDC associated-5 CDC2-like CDC CDK2 CDKI-1C (= P57 or Kip2) MCMs: 3, 4, 5, 6, 7, 8 MCM3APAS (anti-sense RNA) Myc, Max Activator of S-phase kinase (ASK) MAD2 MARK3 HDAC2 M-phase phosphoprotein 6 Origin recognition complexes: 1, 6 RCCs: 1, 2 Thymidine kinase 1

Online Supplementary Table S4. List of genes involved in various “Ingenuity” canonical pathways, whose expression was modified in hematopoietic cells exposed to HOXB4/C4. Ingenuity Canonical Pathways Molecules Role of BRCA1 in DNA Damage Response MLH1, RAD51, SMARCA4, RFC3, BRCA2, RBBP8, TP53, ATM, RFC2 Hypoxia Signaling in the Cardiovascular System LDHA, NFKBIA, UBE2D3, UBE2N, UBE2M, HSP90AA1, UBE2I, HSP90AB1, TP53, ATM, PTEN ATM Signaling RAD51, NFKBIA, SMC1A, TP53, ATM, SMC2, CDK2, SMC3 Cell Cycle: G1/S Checkpoint Regulation MAX, HDAC2, TFDP1, MYC, CCND2, CCNE2, TP53, ATM, CDK2 Role of CHK Proteins in Cell Cycle Checkpoint Control RFC3, TP53, ATM, RFC2, CDK2 Cell Cycle: G2/M DNA Damage Checkpoint Regulation YWHAE, YWHAZ, PRKDC, TOP2B, TP53, ATM EIF2 Signaling GRB2, EIF2AK2, PDPK1, EIF3C, PIK3C2A, EIF3H, EIF3A, EIF4G1, EIF2B5, PPP1CA, EIF2S3, EIF3F, EIF2S1 RAN Signaling RCC1 (includes EG:1104), RANBP1, CSE1L Cell Cycle Regulation by BTG Family Proteins PRMT1, CCNE2, PPP2R1A, CDK2 p53 Signaling PRKDC, STAG1, PIK3C2A, CCND2, TP53, ATM, PTEN, CDK2, PLAGL1 Myc Mediated Apoptosis Signaling GRB2, YWHAE, MYC, YWHAZ, PIK3C2A, TP53 Mitotic Roles of Polo-Like Kinase KIF23, SMC1A, HSP90AA1, CDC20, HSP90AB1, PPP2R1A IGF-1 Signaling IGFBP2, GRB2, YWHAE, PRKCH, PDPK1, YWHAZ, PRKAR1A, PIK3C2A, CSNK2A1 14-3-3-mediated Signaling TUBB, GRB2, YWHAE, PRKCH, YWHAZ, TUBA1A, PIK3C2A, TUBA4A, VIM, TUBB2A Angiopoietin Signaling GRB2, NFKBIA, CRK, TEK, PIK3C2A, ANGPT1 PI3K/AKT Signaling GRB2, YWHAE, NFKBIA, PDPK1, YWHAZ, HSP90AA1, HSP90AB1, PPP2R1A, TP53, PTEN Protein Ubiquitination Pathway USP22, UBE2D3, UCHL5, CDC20, PSMD3, HSP90AA1, UBE2I, USP10, PSMD12, USP14, HSPA8, UBE2M, UBE2N, PSMD11 Online Supplementary Table S5. Genes of the heat shock protein (HSP) family up-regulated by HOXB4/C4 in hematopoietic cells. HSP40 A member 1 A member 2 B member 11 C member 7 HSP70 protein 1A (= HSPA1A, see Table 2) protein 8 HSP70 homolog HSP90 αA1 αB1 β1 (TRA1) HSP105-110 (HSP H1) HSP70/97 organizing protein (STIP1) HSP60 pseudogene1 pseudogene3

Online Supplementary Table S6. Main functions of the genes listed in Table 1A and Table 2 (underlined genes) (from the Gene Database of The National Center for Biotechnology Information (NCBI). EGR1 is an early growth response gene that displays FOS-like functions in several cell types, particularly in hematopoietic cells. It acts as a cell cycle and transcriptional regulator following mitogenic stimulation and as a tumor suppressor gene. MEF2C encodes a transcription factor mainly involved in muscle differentiation, but also promotes myeloid progenitor proliferation as a target of the myeloid-specific miRNA-223 known to negatively regulate progenitor proliferation and granulocytic differentiation. NCL (Nucleolin) is a nucleolar phosphoprotein involved in the synthesis and maturation of ribosomes. TRIM28 mediates transcriptional control by interacting with the Krüppel-associated box repression domain found in many transcription factors. It acts as a retroviral silencer in embryonic stem cells. BPTF is thought to be a transcription regulator. The protein contains a C-terminal bromodomain characteristic of proteins that regulate transcription during proliferation. It is necessary for the chromatin remodeling that occurs during gene transcription. TES is similar to mouse Testin, a testosterone-responsive gene containing 3 LIM domains that mediate protein interactions between transcription factors, cytoskeletal proteins, and signaling proteins. KLF10 (Krüppel-like factor 10) acts as an inducer or repressor of gene transcription to enhance the TGFβ/Smad pathway and other signaling pathways to regulate cell proliferation, differentiation, and apoptosis. GATA3 encodes an enhancer-binding zinc-finger protein that plays important roles during fetal liver hematopoiesis; GATA-3 is also a regulator of T-cell development. HNRPDL binds to pre-mRNAs and plays a role in mRNA splicing and nuclear export. Its expression is down-regulated in leukemic cells induced to proliferate and differentiate with interleukin-4 plus phorbol ester. HNRPK encodes a nuclear RNA-binding protein that complexes with heterogeneous nuclear RNAs (hnRNAs). It influences pre-mRNA processing, metabolism, and transport, and strongly binds to poly(C). This protein is thought to have a role during cell cycle progression. HOXA9 is a member of the Hox complex gene family alike Hoxb4 and Hoxc4, and seems to be involved in hematopoietic cell differentiation. It is frequently over-expressed in human myeloid leukemias; Hoxa9-NUP98 gene fusions are observed in recurrent t(7;11) translocations in some acute myeloid leukemias of the FAB M2 and M4 types. RNPS1 encodes a protein that is part of a postsplicing multiprotein complex involved in mRNA nuclear export and surveillance and that acts by triggering the degradation of mRNAs containing premature stop codons. This protein remains bound to mRNAs after nuclear export, thus acting a nucleo-cytoplasmic shuttling protein. CEBPB encodes a transcription factor involved in regeneration and metabolism of stem cells of numerous tissues, particularly the hematopoietic tissue. TSC22D1 encodes a transcription factor that belongs to the early response gene family and has a transcriptional repressor activity. ZMYND11 encodes a nuclear zinc-finger protein first identified by its ability to bind the adenovirus E1A protein. It is a transcriptional repressor whose activity is inhibited by E1A. HBP1 encodes a transcriptional repressor and cell cycle inhibitor. It is involved in the development of thymocytes, and acts as a tumor suppressor and differentiation inducer of myeloid cells. It also regulates the cell cycle during liver regeneration and has a role in tissue maintenance. Finally, it is known to be a suppressor of Wnt signaling. IKZF1 (ikaros) encodes a key factor for overall lymphocyte development. Mutations of this gene can lead to lymphoid malignancies. ikaros-null mice display pleiotropic hematopoietic cell defects, including stem cell and lymphoid defects besides abnormalities of red cell and megakaryocyte progenitors. LYAR encodes a nucleolar zinc-finger protein that plays a critical role in maintaining embryonic stem (ES) cell identity and growth, and in preventing their apoptosis. The protein forms a complex with nucleolin (NCL, see above) and prevents self-cleavage of that factor which is crucial for ES cell growth. Thus, LYAR controls the stability of NCL which in turn, acts by maintaining ES cell self-renewal. RNF10 encodes a protein containing a ring-finger motif known to be involved in protein-protein interactions. It regulates the expression of myelin-associated glycoprotein, but further specific functions of this protein remain to be determined. SUPT16H is a suppressor of Ty16 element homologs (S. cerevisiae). This actor could play a role through interactions with histones. SYF2 encodes a nuclear protein that interacts with cyclin-D-type binding-protein 1 which is involved in cell cycle regulation at the G1/S transition. Online Supplementary Table S7. Main functions of the tint genes listed in Table 2 (from the Gene Database of The National Center for Biotechnology Information (NCBI). HSPA1A encodes the Hsp70 heat-shock chaperone protein that, among several functions, acts during erythroid differentiation. It interacts with GATA1 in the nucleus of erythroid precursors undergoing terminal differentiation and protects GATA1 from caspase-3 mediated cleavage and proteolysis. DACH1 encodes an ubiquist protein that regulates early progenitor cell proliferation during retinogenesis and pituitary development by directly repressing cyclin-dependent kinase inhibitors. It could also play a role in other tissue stem cell expansion; loss of DACH1 is associated with some poor-prognosis breast cancers and oncogene-dependent tumor metastasis. CDKN1C encodes the cyclin-dependent kinase inhibitor KIP2 that is involved in cell proliferation and differentiation of various tissues. Several mutations of this gene can cause the Beckwith-Wiedemann syndrome, characterized by excessive growth, organomegaly, and tumors. CD53 encodes a protein belonging to the tetraspanin family. This proteins plays a role in the regulation of cell development, activation, growth and motility. This cell surface glycoprotein is known to complex with integrins. Familial deficiency of this gene has been linked to an immunodeficiency associated with recurrent infectious diseases caused by bacteria, fungi and viruses. LGALS3 encodes a member of the galectin family of carbohydrate binding proteins. This protein plays a role in numerous cellular functions including apoptosis, innate immunity, cell adhesion and T-cell regulation. TPM4 encodes a member of the tropomyosin family of actin-binding proteins involved in the contractile system of striated and smooth muscles and the cytoskeleton of non-muscle cells. SOD2 encodes a member of the iron/manganese superoxide dismutase family. This mitochondrial protein binds to the superoxide byproducts of oxidative phosphorylation and converts them to hydrogen peroxide and diatomic oxygen. Mutations in this gene have been associated with idiopathic cardiomyopathy (IDC), premature aging, sporadic motor neuron disease, and cancer. WDR77 encodes a protein of the 20S PRMT5 containing methyltransferase complex, which modifies specific arginines to dimethylarginines in several spliceosomal Sm proteins. This modification targets Sm proteins to the survival of motor neurons (SMN) complex for assembly into small nuclear ribonucleoprotein core particles.

Online Supplementary Figure S1. Expansion of human CD34+ cells cultured in the presence of HOXB4, HOXC4, or both HOXB4 and HOXC4 proteins. Left-side panels represent relative expansion rates of cells exposed to either HOXB4, or HOXC4, or HOXB4 and HOXC4; the 100% reference line corresponds to control expansion in the presence of MS-5/GFP. Right-side pan- els represent absolute cell expansion levels, relative to the day-0 cell input. Online Supplementary Figure S2. Strategy used for transcrip- tome analysis. Duplicate, both-sense experiments were per- formed for either direct (MS-5/HOXB4 versus MS-5/HOXC4) or indirect (MS-5/HOXB4 versus MS-5/GFP + MS-5/HOXC4 ver- sus MS-5/GFP) comparative transcriptome analysis.

Online Supplementary Figure S3. Validation of the transcriptome. RT-qPCR was used to confirm the changes in gene expression observed in the micro-array study. Gene expression changes were measured by duplicate analysis of four RNA samples from purified CD34+ cells co-cultured with MS-5/GFP, MS-5/HOXB4, or MS-5/HOXC4 for 24 h. Relative differences in gene expression were calculated by using the 2-ΔΔCT method, which involves normalizing the CT value for each gene to the CT value of the β2-microglobulin housekeeping gene. Values are shown as the fold induction rela- tive to GFP (mean ± standard deviation). A

B

C

Online Supplementary Figure S4. Main cardinal functions and pathways altered after cell exposure to HOXB4/C4. (A) log(P-value) variation of the expression of genes involved in cardinal cell functions. (B) log(P- value) variation of the expression of main gene pathways deregulated by HOXB4/C4. (C) Proportion of the number of deregulated genes relative to the total number of genes reported for every pathway. All data are derived from Ingenuity software analysis.